Bitcoin

Bitcoin Miners Have Spent $5.11 Billion on AI and Earned $341 Million Back

BlocksBridge put nine public miners at $5.11 billion of capital spending against $341.2 million of AI and HPC revenue in the first half. The revenue side is growing fast, but from a base that is a fraction of the outlay.

⏱ 3 min read Bitcoin
Quick Summary
  • Nine Bitcoin miners spent $5.11 billion on capex in H1 2026 but generated only $341.2 million in AI and HPC revenue, a roughly 15-to-1 ratio
  • AI and HPC revenue accelerated 52% quarter-on-quarter to $205.8 million in Q2 2026, with Core Scientific, TeraWulf and Bitdeer among the leaders
  • BlocksBridge warned that converting mining sites into AI-ready data centres requires substations, cooling systems, networking gear and in some cases GPUs, driving the heavy upfront spend

Public Bitcoin miners are committing billions of dollars to artificial intelligence and high-performance computing infrastructure, but the revenue those investments generate has barely registered compared to the capital flowing out, according to new data from BlocksBridge Consulting.

The numbers behind the gap

In its latest Miner Weekly newsletter, BlocksBridge reported that a combined group of 15 Bitcoin miners and AI data-centre companies spent $30.7 billion on capital assets across their most recent 2026 reporting periods. That figure already exceeds the $21.53 billion the same group spent across all of 2025, a jump of 42.6%.

Narrowing the lens to nine comparable Bitcoin miners, the disparity becomes stark. Those companies deployed $5.11 billion in capital expenditures during the first half of 2026 while generating just $341.2 million in directly reported AI and HPC revenue, producing a capex-to-revenue ratio of roughly 15-to-1.

BlocksBridge calculated capital spending using cash purchases and allocations to hardware, property, equipment and other productive assets, net of proceeds and refunds from asset sales.

AI revenue accelerating, but from a low base

Despite the lopsided ratio, AI and HPC revenue is growing quickly. The same nine miners produced $205.8 million from those operations in the second quarter of 2026 alone, up 52% quarter-on-quarter. Core Scientific, TeraWulf and Bitdeer were among the companies reporting gains during the period.

The steep cost of pivoting to AI

AI and data centres have been positioned as a diversification lifeline for miners squeezed by challenging conditions in the Bitcoin mining sector, but BlocksBridge cautioned that the transition demands far more than existing infrastructure can provide.

‘Power contracts and available land may give miners a starting advantage, but converting those assets into AI-ready capacity requires substations, buildings, cooling systems, networking equipment and, in some business models, GPUs,’ BlocksBridge said.

Bitcoin price provides some relief

Bitcoin surged more than 13% in the week to August 20, climbing back above $72,000 after the US Treasury announced it would at least double the maximum size of its long-term bond buybacks to $4 billion per operation. The move was aimed at improving liquidity in the Treasury market and initially pushed yields lower while lifting risk appetite.

Whether that recovery translates into relief for companies still running large-scale Bitcoin operations remains to be seen.

CoinShares broadens its miner ETF universe

In a sign of how far the mining sector has shifted toward AI and HPC, CoinShares announced a strategy change for its industry-tracking exchange-traded fund. Now rebranded as the CoinShares Bitcoin Mining and Digital Power ETF under the ticker WGMI, the fund holds $222.4 million in assets under management and encompasses 29 holdings drawn from Bitcoin miners, data centre operators, AI semiconductors, power generation firms and HPC companies. CoinShares describes the universe as ‘the businesses powering the digital economy.’

⚖️ Our Verdict ⚖️ Watch and Wait

Fifteen dollars of capital spending for every dollar of AI revenue is a wide gap, and the build-out costs BlocksBridge lists are the kind that arrive before any customer does. But the revenue side grew 52% in a single quarter, which is what an early-stage business looks like rather than a failing one, and the question is simply whether it keeps compounding fast enough to catch up with what has already been spent.